I'm suspicious of the validity of these studies if their metric was standardized testing
Standardized tests don't measure everything. But if a kid does poorly on a standardized math test, that most likely means the kid didn't learn math. If a student can't read a paragraph and answer questions about it, it most likely means the student can't read well. Reducing class sizes is expensive, more so than almost any other educational intervention. It is not acceptable to assume that it "just works" in the absence of evidence.
Educational reform has a long history of "faddism", where changes are made in the absence of evidence. "New math" and "Whole Word Learning" were inflicted on millions of kids before their folly was realized. But in those cases we could fix the problem by just replacing the textbooks. Smaller classrooms often involve structural changes to the school. That will be much harder to fix if it turns out to be yet another mistake.
Also, how do you conclude that there's no benefit, and then go on to explain the real reason for those nonexistent benefits?
Nobody said it doesn't work. Just that it doesn't work well enough or broadly enough to justify the cost. What we should be doing is identifying the situations, and the particular children that most benefit. This appears to be young (K-3) children with disadvantaged backgrounds. Currently this is the opposite of what we do. Small classrooms are mostly commonly used in high-income areas with brighter kids that get little benefit.